Interpretation of a lagged dependent variable, coefficient as proportion- Panel data

Dear experts!
I am having some difficulties in interpreting the results of my regression model using Blundell–Bond linear dynamic panel-data estimation (xtdpdsys).

Countries in scope: 7, firms: 30, time period 15 years (annual observations)

My dependent variable is:
- RD expenditure/total assets

L1= Lagged 1 year; D1= differenced
RD: Research & Development expenditure

Independent variables:
- L1. RD expenditure/total assets (coefficient: 0.66)
- L1. cash flow/total assets
- L1. Debt/total assets
- Revenue/total assets (coefficient 0.04)
- D1. Revenue/total assets (coefficient 0.06)
- Log of employees
- A macroeconomic variable (% of GDP)/1000 (coefficient 0.26)

I will highlight the significant coefficients, which would be great if anyone have time to help.

The lagged dependent varaibles gives me a coefficient of 0.66 - What does this exactly mean?

Furthermore, Revenue/total assets give a coefficient of 0.04 and the differenced (revenue/total assets) gives 0.06. How can I interpret this?

Lastly, the macroeconomic variable provides a coeffcient of 0.26

I have included respective year dummies where 5 out of 15 are showing significance, (coefficients ranging from 0.006-0.010, are there any way to put interpretation on this?

I am crossing my fingers and hope to hear from you!

Best regards,